Recent work in forgetting and spacing of learning across long time scales suggests that which has been learned is easily forgotten. In order to be able to prevent this forgetting from happening in educational contexts, it is useful to understand how this forgetting occurs outside of the laboratory. Analyses are conducted of the time course of forgetting for declarative fact and Mathematics learning from computer assisted instruction systems, in order to see how forgetting occurs in ecologically valid contexts. The difference between the learning of declarative facts and Mathematics skills is compared and contrasted, showing that individual Mathematics questions behave similarly to fact learning, but that the learning of Mathematics skills may potentially have differing long term trajectories.

The computations that link our perception to action, human cognition, define a multifaceted, dynamic and complex system. In the center is a body housing a brain that develops and evolves over our lifespan. Ever unfolding processes towards individuality are guided by temporally and spatially varying expression of inherited programs, pushed and prodded by experience and embedded in rich cultural contexts. One’s journey through this landscape injects individuality into cognition and may be observable in our actions. Because cognition is latent, individual differences are amorphous, taking refuge in factors localized or distributed. Despite this, for over a century we have attempted to characterize the relative importance of inherited factors in cognitive individuality. Often cognitive differences are operationalized, particularly those thought to emerge from internal factors, as variability in neuropsychological performance or in the presence of psychiatric diagnoses. Although these instruments capture restricted aspects of human cognition, genetic analyses suggest individuals sharing more inherited factors index more similarly on these measures. To me, this recurrent trend suggests cognitive systems involving genetically similar individuals integrate similar percepts into more similar actions. Insights into the nature of these genetic factors becomes an interesting aim for cognitive scientists interested in parameters that may define human cognition.

This thesis is a series of works that attempt to make concrete an approach for characterizing features of genetic factors involved in aspects of human cognition. I focus on data resources from genome-wide association studies (GWAS). GWAS make available surveys of the involvement of millions of common single nucleotide polymorphisms (SNPs) in numerous neuropsychological and psychiatric metrics for hundreds of thousands of individuals. These data were rapidly assembled assuming one particular analytic framework and are controversial with regards their success. Here I follow emerging trends that conceptually challenge this original design, suggesting alternative approaches might increase the utility of these data. Specifically, I propose that the traditional approach is biased a lens that does not allow for many thousands of genetic factors underlying cognitive traits. Readjusting our view of existing data resources uncovers new insights and demonstrate a potential for polygenic studies of human cognition.

Children born before 33 weeks of gestation have an increased risk for early brain injury, as well as cognitive, behavioral, and academic deficits. The aim of this dissertation was to contribute to a more comprehensive understanding of brain and cognitive development following preterm birth. All three studies used data from a cohort of children recruited before starting kindergarten (5 years of age) who were followed up after one and two years. At each timepoint, children completed structural and diffusion weighted MRI, as well as a battery of cognitive and behavioral tests. Forty-seven to 51 children born preterm (24–32 weeks gestational age), and 27 to 28 children born full-term were included in each study.The first study examined the effect of preterm birth on number and arithmetic skills, and how it is mediated by related cognitive functions. Number and arithmetic skills were lower in children born preterm. The performance gap in number skills decreased over time, while the performance gap in arithmetic skills increased. The effect of preterm birth on number and arithmetic skills was mediated by phonological processing, visual-motor integration, and inhibitory control, but not spatial working memory. Phonological processing showed the strongest mediating effect. The second study examined the effect of preterm birth on development of subcortical gray matter and white matter volumes, and diffusivity measures of white matter tracts. Children born preterm had smaller volumes of thalamus, brain stem, cerebellar white matter, cingulum, corticospinal tract, inferior frontal occipital fasciculus, uncinate fasciculus, and temporal superior longitudinal fasciculus, while their ventricles were larger compared to full-term controls. We found no significant effect of preterm birth on diffusivity measures. Despite developmental changes and growth, group differences were present and similarly strong at 5, 6, and 7 years. The third study examined the association between white matter tract diffusivity measures and mathematics skills at 5 and 7 years of age. Fractional anisotropy of the right and left corticospinal tract, left inferior longitudinal fasciculus, and left inferior frontal occipital fasciculus showed a significant interaction effect with term status. This moderating effect of preterm birth may be indicative of reorganization and plasticity of functional networks following early injury due to prematurity. These findings help to delineate the developmental trajectory of brain and cognitive development during early childhood following preterm birth. This knowledge may provide guidance for opportunities for support of children born preterm and their families, such as targeted evaluation and use of interventions.

The acquisition of a second language often occurs during early childhood, a period of dynamic change in cognitive, language, and brain development.  This thesis examined the contribution of two important factors in bilingualism, language experience and SES, in association with individual differences in language skills, executive function, and white matter maturation in bilingual children.  Language skills in both English and Spanish were measured, including vocabulary, verbal fluency, and spoken narrative skills.  Executive function measures for response inhibition, inhibitory control, and cognitive flexibility were also examined. While bilinguals had lower receptive English vocabulary than the monolingual children, within the bilingual group, language experience was a contributing factor to language outcomes in both languages.  When both SES and language experience were entered as predictors of the language measures, SES was only significant for English expressive vocabulary scores in bilinguals.  Bilingual children underperformed on executive function tasks compared to the monolingual children, although, SES accounted for group differences in cognitive flexibility (though not inhibitory control) performance.  The group comparisons revealed an interaction effect of language group by SES for response inhibition, such that among the bilingual children, there was a negative association between SES and performance, but not for the monolingual children.  Within the bilingual group, language experience and SES both were contributing factors to individual differences in executive function performance, while differing patterns of associations for executive function measures emerged.  In addition, white matter maturation was compared in bilingual children to monolingual children in three white matter fiber tracts, the inferior fronto-occipital fasciculus (IFOF), the superior longitudinal fasciculus (SLF), and the anterior cingulum.  Monolingual children showed greater white matter maturation in the SLF compared to the bilingual group, independent of SES.  Post-hoc analysis revealed group differences in white matter maturation in the separate segments of the IFOF, such that monolinguals had more mature values on white matter measures in the anterior and middle IFOF, while bilinguals had more mature values in the posterior IFOF.  Within the bilingual group, language experience was related to more mature white matter characteristics in the anterior cingulum, a brain region thought to support language control in bilinguals.

Human development is a dynamic, protracted process influenced by genetics, the environment, experience, and many other factors. During development, an individual develops an entirely unique neural architecture along with a unique set of cognitive skills and abilities. Through recent large-scale pediatric neuroimaging initiatives, we now have a better understanding of the protracted structural changes occurring in the developing brain. We are also beginning to map out relationships between developing brain structure and domain-specific cognition, although we haven’t fully characterized these relationships and their variability throughout typical development. The goal of this work is to assess how individual differences in regional cortical and subcortical brain structure are related to behavioral and cognitive variability in different domains during childhood and adolescence, which is a period of rapid dynamic change. This work also aims to investigate overlapping as well as distinguishing characteristics of the relationships between developing neural systems and domain-specific cognition.

In this dissertation, I focus on three cognitive domains: phonological awareness, spatial working memory, and response inhibition. These three cognitive domains are thought to involve relatively distinct brain regions and networks, allowing us to investigate the specificity of associations between structure and function in the developing brain. In addition, these three domains have been well studied in pediatric, clinical, and adult populations in behavioral and functional neuroimaging studies, leading to relatively well-defined region-specific hypotheses. Utilizing three distinct cognitive domains also supports the investigation of domain-general aspects of cognitive development, such as latent factors or skills supporting multiple areas of cognitive development. In addition, studying multiple cognitive domains allows me to study the reflection of any shared features in the neural architecture. I will first address the region-specific associations between cortical and white matter regions thought to be principally related to each cognitive domain. Following that, I will carry out a more data-driven analysis aimed at exploring possible latent factors underlying the associations between these cognitive domains and structural regions. These results may provide insight into the neurobiological correlates of cognitive development and the nature of individual difference variability during development.